Lightspeed’s Anthropic Megadeal worth $ 2 billion reinforces the AI ​​ambitions of VC Firm

Just like a large part of the VC industry, Lightspeed has shifted his attention to AI startups, in which companies are supported at an early stage, such as the music company Suno Inc. And the video startup Pika, next to larger players. In December it said goodbye to his two most important consumer investors and said that it was adjusting his consumer investment strategy to better match the ‘Ai era’.

In total, Lightspeed has already invested 2.2 billion dollars in AI-Deals, a figure in which the last Anthropic investment is not included, according to someone else who is familiar with the case. Soon it will have extra firepower to be able to attack the money eager for money. The end is approaching a fundraising that is expected to yield seven billion dollars, a person who was familiar with the case said. A spokesperson for Lightspeed refused to comment on fundraising. The information was previously reported on the fundraising efforts.

The company’s anthropic investment is one of the most ambitious so far. And although the value of $ 60 billion appears to be high, the Lightspeed partners have good hope that the deal will ever look like a bargain.

“Overall, it feels like the ratings are expensive, because we see a lot of activity and close many deals,” said Lightspeed partner Guru Chahal last year at a Fortune Brainstorm Tech conference. “If you look back, every round seemed incredibly expensive at the time and in retrospect, incredibly cheap.”

Large AI deals remain a source of discussion in Silicon Valley. Although the largest companies appear to be the most transforming, some venture capitalists claim that participation in enormous financing rounds will not yield the return that technology investors need to satisfy their supporters. These investors focus on smaller AI apps and services, rather than the giants such as Anthropic and OpenAi, who are involved in the development of the expensive building blocks of industry.

The recent proliferation of AI-Megadeals also indicates a broader shift in risk capital: a deviation from the traditional strategy of investments at an early stage, in which companies acquire greater interests against lower valuations. Now venture capital companies pay a large premium and bet that a small number of AI companies could eventually be worth more than 1 trillion dollar.

The growing size of venture capital funds has also forced companies to write out larger checks, says Weber. Instead of striving for enormous multiplication of their investments, companies try to “not necessarily find homeruns, they try to find ways to double their money,” he said.

“Nowadays there are only a limited number of iconic pre-ipo companies from generation to generation,” says IVP General Partner Ajay Vashee. “If it is your mandate to invest at that stage, then you must find opportunities to put your capital to work.”

Shaky start

The race to find those opportunities is loaded with risks, including uncertainty about the regulations, severe competition and rising infrastructure costs for leading AI developers.

Investors fear that their AI bets are failing, which puts companies in danger when the bubble is bursting. The sector has already seen several billions of companies stumble.

Lightspeed, for example, was a co-leader of a controversial investment in Stability AI, the developer of the image generator Stable Diffusion, which was appreciated at 1 billion dollars in 2022. Shortly thereafter, several important developers resign from the company in the midst of rising tensions with the whimsical chief executive. Agent Emad Mostaque, lawsuits and financial problems. Mostaque resigned from the company in early 2024. The company has since appointed a new CEO and collected extra capital, reports Bloomberg.

Lightspeed is also an important investor in Mistral, the Paris-based open-source company that is now competing with a whole series of better funded language models.

Of course Lightspeed and other prominent VC companies have good hope that the placement of several bets on competing companies will yield at least one large AI winner. If that is not the case, the consequences can be considerable.

“You can’t lose too many games with this poker with high bet,” said Guleri of Sierra Ventures. “That is the risk of the strategy.”

-with help from Katie Roof.

Most read from Bloomberg Businessweek

© 2025 Bloomberglp